February 2003 - Test - lists.nanog.org

Re: Symantec detected Slammer worm "hours" before
by hostmaster 13 Feb '03

13 Feb '03

DeepSight is SecurityFocus. Their claim may have some truth in it. But, so does the 19000+ partners. They mean customers....but not necessarily customers/ subscribers to DeepSight. (they may have 'accidentally' included all their SecurityFocus lists' subscribers in that number as well :). Though I do appreciate the SF team, I haste to add that this revelation smells seriously commercially tainted. All of the major networks would see such anomaly pretty fast, and at virtually the same time. This is however the right moment to underscore 'corporate fears', and bring attention to their relatively new service/product. cheers Bert director research nso/ft. myers

1 0

RE: Voice over IP - performance
by Bender, Andrew 13 Feb '03

13 Feb '03

Don't forget that the signaling agents that drive the DSPs also contribute to load on the host / control CPU. We have found that this can be a very willing consumer of utilization on the platforms under discussion... folks with super low hold calls would be the ones likely to be challenged by this; particuarly with those boxes that still have the 4500-class processor in them. Regards, Andrew taqua.com > -----Original Message----- > From: Mathew Lodge [mailto:Mathew@CPlane.com] > Sent: Wednesday, February 12, 2003 6:46 PM > To: Charles Youse; Bill Woodcock > Cc: nanog(a)nanog.org > Subject: RE: Voice over IP - performance > > > > At 03:23 PM 2/12/2003 -0500, Charles Youse wrote: > >I'm assuming in the case of, e.g., a 2650 + dual T-1 PRI > interface can > >actually encode/decode 48 simultaneous g729a voice streams without > >issues? Any idea what the CPU utilisation is - or is this > handled in > >separate DSPs in the voice network module itself? > > On these particular Cisco boxes, the DSP does the all audio > filtering, > CODEC functions, echo cancellation, jitter buffering & > adjustment, silence > suppression (AKA voice activity detection, if you turned it > on), and also > prepends the RTP and IP headers. The router CPU just has to > forward the > packet that's generated by the DSP. > > Router CPU utilization is therefore a function of the number > of packets per > second that the voice card generates and the size of each > packet, plus > signaling overhead. The packet size and rate depend on the > CODEC itself > (higher compression CODECs generate smaller packets), the > sample size (20ms > is the Cisco default, reducing or increasing it makes the > packets smaller > or larger and the packet rate higher or lower, respectively), > and whether > voice activity detection is on (roughly halves the packet rate). > > If you leave the default settings in place (no VAD, 20ms > sample size), > you'll be OK with any of the CODECs. > > Mathew. > > > >C. > > > >-----Original Message----- > >From: Bill Woodcock [mailto:woody@pch.net] > >Sent: Wednesday, February 12, 2003 2:43 PM > >To: Charles Youse > >Cc: nanog(a)nanog.org > >Subject: Re: Voice over IP - performance > > > > > > > Does anyone have any real-world figures for VoIP > performance on > > > various platforms? In other words, how many calls > can an otherwise > > > unused e.g., Cisco 2600 be expected to handle if it's > the conversion > > > point from trunked voice calls to IP. Some rough numbers for > > > different codecs on different hardware would be very > useful. Most > > > specifically I'm interested in Cisco router platforms > but other > > > vendor stats would be appreciated as well. > > > >Actually I just ran the dollars-per-simultaneous-call numbers for > >different models for some friends. I'll append it. > Basically, if you run > >g711, you're limited by the number of PRI channels on the > box. If you run > >g729a, you're limited by the number of DSPs you can fit in > the box. The > >numbers I ran were assuming g729a. > > > > -Bill > > > > > > > > Cost > > per > >Package which can handle 23 simultaneous calls: call > >CISCO1760 10/100 Modular Router $1,595 > >VWIC-1MFT-T1 1-Port RJ-48 Multiflex T1 $1,300 > >PVDM-256K-12 3-DSP Module (9 calls) $1,200 > >PVDM-256K-20HD 5-DSP Module (15 calls) $4,000 > >Total $8,095 $352 > > > >Different package which can handle 23 simultaneous calls: > >CISCO2650 10/100 Modular Router $3,295 > >NM-HDV-1T1-24E Single-Port T1 Voice NM $9,100 > >Total $12,395 $539 > > > >Package which can handle 45 simultaneous calls: > >CISCO2650 10/100 Modular Router $3,295 > >NM-HDV-2T1-48 Dual-Port T1 Voice NM $9,800 > >Total $13,095 $291 > > > >Package which can handle 46 simultaneous calls: > >CISCO2650 10/100 Modular Router $3,295 > >NM-HDV-2T1-48 Dual-Port T1 Voice NM $9,800 > >PVDM-256K-20HD 5-DSP Module (15 calls) $4,000 > >Total $17,095 $372 > > > >Upgradeable package which can handle 46 simultaneous calls: > >AS535-2T1-48-AC-V AS5350-V/2T1 $18,900 $411 > > > >Package which can handle 92 simultaneous calls: > >AS535-4T1-96-AC-V AS5350-V/4T1 $33,600 $366 > > > >Package which can handle 184 simultaneous calls: > >AS535-8T1-192-AC-V AS5350-V/8T1 $58,700 $319 > > > >Upgradeable package which can handle 184 simultaneous calls: > >AS54HPX-8T1-192AC AS5400HPX/8T1 $65,500 $356 > > > >Package which can handle 644 simultaneous calls: > >AS54HPX-CT3-648AC AS5400HPX/CT3 $170,300 $265 > >

1 0

New IANA IPv4 allocations and bogon updates
by Rob Thomas 13 Feb '03

13 Feb '03

Hi, NANOGers. As announced by John Crain earlier today, IANA has allocated 222/8 and 223/8 to APNIC as of February 2003. The sundry bogon projects have been updated to reflect these allocations. You will find the catch-all Bogon Reference Page here: http://www.cymru.com/Bogons/index.html This includes links to the Secure IOS Template, Secure Juniper Template, bogon monitoring, and other associated references. Both 222/8 and 223/8 were withdrawn from the bogon route-server as of 13 FEB 02:01:00 US CST. Thanks! Rob, for Team Cymru. -- Rob Thomas http://www.cymru.com ASSERT(coffee != empty);

1 0

IPv6 subnet calculators?
by Jeffrey Wheat 13 Feb '03

13 Feb '03

I am trying to find an IPv6 subnet calculator or even a "cheat sheet" that will help show how a /32 allocation is broken down into /40, /48 and /64 subnets. Has anyone found such a tool? I have been using the NorthStar web based database tool for dealing with my IPv4 allocations. While the author has mentioned support for IPv6 in the future, I am trying to get things rolling now. Any information is greatly appreciated. Best regards, Jeff --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com) Version: 6.0.449 / Virus Database: 251 - Release Date: 1/27/2003

3 2

RE: IPv6 subnet calculators?
by Irwin Lazar 13 Feb '03

13 Feb '03

Try: http://freshmeat.net/projects/sipcalc/?topic_id=150 > -----Original Message----- > From: Jeffrey Wheat [mailto:jeff@cetlink.net] > Sent: Wednesday, February 12, 2003 5:58 PM > To: nanog(a)merit.edu > Subject: IPv6 subnet calculators? > > > > I am trying to find an IPv6 subnet calculator or even a > "cheat sheet" that > will help show how a /32 allocation is broken down into /40, > /48 and /64 > subnets. Has anyone found such a tool? I have been using the NorthStar > web based database tool for dealing with my IPv4 allocations. > While the > author has mentioned support for IPv6 in the future, I am > trying to get > things rolling now. Any information is greatly appreciated. > > Best regards, > Jeff > > --- > Outgoing mail is certified Virus Free. > Checked by AVG anti-virus system (http://www.grisoft.com) > Version: 6.0.449 / Virus Database: 251 - Release Date: 1/27/2003 > >

2 1

Re: [Re: Cisco VoIP/dominatrix link explored]
by Joshua Smith 12 Feb '03

12 Feb '03

Bill Woodcock <woody(a)pch.net> wrote: > > On Wed, 12 Feb 2003, neal r wrote: > > Voice over IP is a stinking, fetid swamp, and I suggest if anyone > > feels compelled to waste money and feel pain while doing something > > with telephones and the internet... > > All I can say is, you must not be doing it right. I know a lot of people, > myself included, who are doing it on a large scale now, and it's > relatively painless, particularly as compared with the PSTN. > > If an existence proof isn't sufficient for you, I'm not quite sure how to > communicate with you. > > -Bill > not that bill needs my approval, but i agree completely. the one thing that i have run into with voip is hardware compatibility - each vendor tells you one thing, and then an engineer who does it in the field tells you something else (or perhaps this is unique to my experience). despite this, i have voip routes running around the world, and they do a decent number of minutes - and to add my $0.02 to the thread from the past two days, most of the traffic is over the public internet, with little (read no) qos in place. if i only have a single e1, i don't push more than an e1 worth of traffic there (no fancy queuing, just simple engineering) "Walk with me through the Universe, And along the way see how all of us are Connected. Feast the eyes of your Soul, On the Love that abounds. In all places at once, seemingly endless, Like your own existence." - Stephen Hawking -

1 0

RE: Voice over IP - performance
by Mathew Lodge 12 Feb '03

12 Feb '03

At 03:23 PM 2/12/2003 -0500, Charles Youse wrote: >I'm assuming in the case of, e.g., a 2650 + dual T-1 PRI interface can >actually encode/decode 48 simultaneous g729a voice streams without >issues? Any idea what the CPU utilisation is - or is this handled in >separate DSPs in the voice network module itself? On these particular Cisco boxes, the DSP does the all audio filtering, CODEC functions, echo cancellation, jitter buffering & adjustment, silence suppression (AKA voice activity detection, if you turned it on), and also prepends the RTP and IP headers. The router CPU just has to forward the packet that's generated by the DSP. Router CPU utilization is therefore a function of the number of packets per second that the voice card generates and the size of each packet, plus signaling overhead. The packet size and rate depend on the CODEC itself (higher compression CODECs generate smaller packets), the sample size (20ms is the Cisco default, reducing or increasing it makes the packets smaller or larger and the packet rate higher or lower, respectively), and whether voice activity detection is on (roughly halves the packet rate). If you leave the default settings in place (no VAD, 20ms sample size), you'll be OK with any of the CODECs. Mathew. >C. > >-----Original Message----- >From: Bill Woodcock [mailto:woody@pch.net] >Sent: Wednesday, February 12, 2003 2:43 PM >To: Charles Youse >Cc: nanog(a)nanog.org >Subject: Re: Voice over IP - performance > > > > Does anyone have any real-world figures for VoIP performance on > > various platforms? In other words, how many calls can an otherwise > > unused e.g., Cisco 2600 be expected to handle if it's the conversion > > point from trunked voice calls to IP. Some rough numbers for > > different codecs on different hardware would be very useful. Most > > specifically I'm interested in Cisco router platforms but other > > vendor stats would be appreciated as well. > >Actually I just ran the dollars-per-simultaneous-call numbers for >different models for some friends. I'll append it. Basically, if you run >g711, you're limited by the number of PRI channels on the box. If you run >g729a, you're limited by the number of DSPs you can fit in the box. The >numbers I ran were assuming g729a. > > -Bill > > > > Cost > per >Package which can handle 23 simultaneous calls: call >CISCO1760 10/100 Modular Router $1,595 >VWIC-1MFT-T1 1-Port RJ-48 Multiflex T1 $1,300 >PVDM-256K-12 3-DSP Module (9 calls) $1,200 >PVDM-256K-20HD 5-DSP Module (15 calls) $4,000 >Total $8,095 $352 > >Different package which can handle 23 simultaneous calls: >CISCO2650 10/100 Modular Router $3,295 >NM-HDV-1T1-24E Single-Port T1 Voice NM $9,100 >Total $12,395 $539 > >Package which can handle 45 simultaneous calls: >CISCO2650 10/100 Modular Router $3,295 >NM-HDV-2T1-48 Dual-Port T1 Voice NM $9,800 >Total $13,095 $291 > >Package which can handle 46 simultaneous calls: >CISCO2650 10/100 Modular Router $3,295 >NM-HDV-2T1-48 Dual-Port T1 Voice NM $9,800 >PVDM-256K-20HD 5-DSP Module (15 calls) $4,000 >Total $17,095 $372 > >Upgradeable package which can handle 46 simultaneous calls: >AS535-2T1-48-AC-V AS5350-V/2T1 $18,900 $411 > >Package which can handle 92 simultaneous calls: >AS535-4T1-96-AC-V AS5350-V/4T1 $33,600 $366 > >Package which can handle 184 simultaneous calls: >AS535-8T1-192-AC-V AS5350-V/8T1 $58,700 $319 > >Upgradeable package which can handle 184 simultaneous calls: >AS54HPX-8T1-192AC AS5400HPX/8T1 $65,500 $356 > >Package which can handle 644 simultaneous calls: >AS54HPX-CT3-648AC AS5400HPX/CT3 $170,300 $265

1 0

Re: Cisco VoIP/dominatrix link explored
by Bill Woodcock 12 Feb '03

12 Feb '03

On Wed, 12 Feb 2003, neal r wrote: > Voice over IP is a stinking, fetid swamp, and I suggest if anyone > feels compelled to waste money and feel pain while doing something > with telephones and the internet... All I can say is, you must not be doing it right. I know a lot of people, myself included, who are doing it on a large scale now, and it's relatively painless, particularly as compared with the PSTN. If an existence proof isn't sufficient for you, I'm not quite sure how to communicate with you. -Bill

1 0

RE: Voice over IP - performance
by Charles Youse 12 Feb '03

12 Feb '03

I'm assuming in the case of, e.g., a 2650 + dual T-1 PRI interface can actually encode/decode 48 simultaneous g729a voice streams without issues? Any idea what the CPU utilisation is - or is this handled in separate DSPs in the voice network module itself? C. -----Original Message----- From: Bill Woodcock [mailto:woody@pch.net] Sent: Wednesday, February 12, 2003 2:43 PM To: Charles Youse Cc: nanog(a)nanog.org Subject: Re: Voice over IP - performance > Does anyone have any real-world figures for VoIP performance on > various platforms? In other words, how many calls can an otherwise > unused e.g., Cisco 2600 be expected to handle if it's the conversion > point from trunked voice calls to IP. Some rough numbers for > different codecs on different hardware would be very useful. Most > specifically I'm interested in Cisco router platforms but other > vendor stats would be appreciated as well. Actually I just ran the dollars-per-simultaneous-call numbers for different models for some friends. I'll append it. Basically, if you run g711, you're limited by the number of PRI channels on the box. If you run g729a, you're limited by the number of DSPs you can fit in the box. The numbers I ran were assuming g729a. -Bill Cost per Package which can handle 23 simultaneous calls: call CISCO1760 10/100 Modular Router $1,595 VWIC-1MFT-T1 1-Port RJ-48 Multiflex T1 $1,300 PVDM-256K-12 3-DSP Module (9 calls) $1,200 PVDM-256K-20HD 5-DSP Module (15 calls) $4,000 Total $8,095 $352 Different package which can handle 23 simultaneous calls: CISCO2650 10/100 Modular Router $3,295 NM-HDV-1T1-24E Single-Port T1 Voice NM $9,100 Total $12,395 $539 Package which can handle 45 simultaneous calls: CISCO2650 10/100 Modular Router $3,295 NM-HDV-2T1-48 Dual-Port T1 Voice NM $9,800 Total $13,095 $291 Package which can handle 46 simultaneous calls: CISCO2650 10/100 Modular Router $3,295 NM-HDV-2T1-48 Dual-Port T1 Voice NM $9,800 PVDM-256K-20HD 5-DSP Module (15 calls) $4,000 Total $17,095 $372 Upgradeable package which can handle 46 simultaneous calls: AS535-2T1-48-AC-V AS5350-V/2T1 $18,900 $411 Package which can handle 92 simultaneous calls: AS535-4T1-96-AC-V AS5350-V/4T1 $33,600 $366 Package which can handle 184 simultaneous calls: AS535-8T1-192-AC-V AS5350-V/8T1 $58,700 $319 Upgradeable package which can handle 184 simultaneous calls: AS54HPX-8T1-192AC AS5400HPX/8T1 $65,500 $356 Package which can handle 644 simultaneous calls: AS54HPX-CT3-648AC AS5400HPX/CT3 $170,300 $265

2 1

Voice over IP - performance
by Charles Youse 12 Feb '03

12 Feb '03

Does anyone have any real-world figures for VoIP performance on various platforms? In other words, how many calls can an otherwise unused e.g., Cisco 2600 be expected to handle if it's the conversion point from trunked voice calls to IP. Some rough numbers for different codecs on different hardware would be very useful. Most specifically I'm interested in Cisco router platforms but other vendor stats would be appreciated as well. Off-list replies please; if anyone is interested in the results I will condense the responses and post back to the list. Thanks, C.

2 1